A Data Imputation Method with Support Vector Machines for Activity-Based Transportation Models

Activity-based approaches in transportation models aim at predicting which activities are conducted where, when, for how long, with whom, the transport mode involved and so on. An activity-based framework named FEATHERS (Forecasting Evolutionary Activity Travel of Households and their Environmental Repercussion S) has been developed for Flanders in Belgium. During the establishment of the framework, lots of data are needed. One of the main data sources are activity-based diaries. However, activity diaries tend to contain incomplete information due to various reasons. More recently, with the development of computer science and technology, some artificial intelligence and machine learning techniques have arisen to process the missing data. In this chapter, a data imputation method with a Support Vector Machine (SVM) is proposed to solve the issue of missing data in activity-based diaries. In order to verify the efficiency of SVMs, other methods such as LDA (Linear Discriminant Analysis) and PNN (Probabilistic Neural Network) are also used to process the same data imputation problem. Compared with accuracies obtained by SVMs, the accuracies obtained by LDA and PNN are lower. The initial experimental results show that missing elements of observed activity diaries can be accurately inferred by relating different pieces of information. Therefore the proposed SVM data imputation method in this chapter serves as an effective data imputation method that can induce complete activity diaries in the case of missing information.